Same Bar, More Than One Door: Assessing Across Cultures

Two finalists for a European group's country-manager role in the Gulf. The first, Milan-based and American-trained, gave interviews the client described as electric: he claimed his achievements in the first person, filled every silence, and left the panel feeling, in their word, led. The second, a German operator with the stronger verified record (better scope audit, better second ring, a cleaner governance file), spoke in measured paragraphs, attributed liberally to his teams, and let three silences stand. The debrief lasted ten minutes. "He didn't fill the room," the client said of the German, and the sentence was meant to end the conversation.

It shouldn't have, and eventually it didn't, because someone at the table asked the question this chapter is built around: filled the room according to whom? The room is not the job. The job was building a business across Saudi and Emirati stakeholders, an environment where, as it happens, neither the American style nor the German one is native, and what the panel had scored as "presence" was, on inspection, proximity to one culture's performance of leadership, measured by assessors formed in that same culture. The record said one thing; the room said another; and the room was measuring itself.

This chapter is the method for not doing that, and, just as firmly, for not overcorrecting into its opposite. Its principle fits in six words, and everything else is execution: same bar, more than one door. The evidence standard never moves. The routes by which a candidate is allowed to reach it multiply. Fail the first discipline and you are scoring conformity to your own culture's style and calling it capability: ethnocentric assessment. Fail the second and you are excusing missing evidence as cultural difference, the soft bigotry of the lowered bar, which insults the candidate and betrays the client at the same time. Both failures are live in every cross-border search, and this chapter polices both.

What actually varies: the evidence

Three bodies of research tell you where cultural pattern genuinely distorts assessment, and they are specific enough to design against.

Self-presentation varies — a lot — and it moves scores. In the key experimental work, self-promotion in Western interview settings is not merely tolerated but expected: European-heritage applicants self-promoted more than East Asian applicants, chronic self-promoters were rated most hirable, and, the finding that should end any complacency, the gap survived when judges read transcripts only, stripped of face and voice. The mechanism wasn't accent prejudice; it was that active ingratiation behaviors, engagement and humor and visible enthusiasm, were more common in one group's interview grammar, and raters read the grammar as merit. The cross-national surveys complete the map: American respondents score strikingly high on intended interview self-presentation; Germans and Norwegians strikingly low. Which means a quietly strong Munich candidate assessed against a Boston-formed comparator starts the meeting with an invisible handicap that has nothing to do with the job, which is precisely our vignette, measured.

The grammar of dissent and communication varies with power distance. Whether truth travels upward loudly or through choreography, whether disagreement happens in the meeting or in the corridor before it, whether directness reads as clarity or as rudeness: these are patterned by region, and the GLOBE program mapped the pattern. Participative leadership is central to the prototype in Germanic Europe and the Anglo world, far less so in the Middle East cluster, with Latin Europe between; self-protective styles are despised in some clusters and tolerated in others. Chapter 12's candidate from Stuttgart, Riyadh, or Central Asia runs the same escalation through different machinery, and an anchor written in one culture's machinery scores the machinery, not the escalation.

But some substance is universal — and this is what saves the method. GLOBE's most useful result for assessors is not the differences; it is the short list of attributes endorsed everywhere: integrity, trustworthiness, vision, performance orientation, being planful, communicative, team-building, and the universally rejected: egocentric, irritable, dictatorial. The universal list is where your bar lives. The contingent list, meaning participative and autonomous and self-protective and humane-oriented styles, is where the doors multiply. An anchor set built on the universal substance travels; an anchor set built on one region's style is a mirror wearing a rubric.

The interview, adjusted

Four adjustments, each preserving the standard it serves.

Brief every candidate on the process, in advance, identically. What competencies will be assessed, what question formats to expect, what a complete answer contains. This feels like giving the game away; it is the opposite. The hidden advantage in unstructured, unexplained interviews belongs to candidates socialized into Anglo interview grammar: they already know the game. Explicit expectations shrink that inherited edge, and the structured-interview evidence says transparency of this kind costs validity nothing.

Elicit episodes down a ladder, not off a cliff. The standard Western prompt, tell me about your biggest achievement, demands first-person claiming as the price of entry, and in modesty-normed cultures the price is unpayable: the capable candidate answers institutionally, the assessor hears evasion, and the score records a norm as a deficit. The ladder fixes it without lowering anything. Start at the team and descend. Describe the situation your team faced. What decisions or interventions were specifically yours? What changed because of what you did? What would have happened without you? By the fourth rung the candidate has made exactly the individual claims the bar requires (ownership, per Chapter 2's four tests, is still mandatory), but the door opened at an altitude their grammar permits. The same ladder, note, is simply good practice on every candidate; it merely stops being optional across cultures.

Write anchors function-first — always, but here above all. The rule every cluster chapter carried now gets its full justification: surfaces difficult information upward, in time, effectively travels across the GLOBE map; openly challenges the boss in the meeting is Frankfurt wearing a scoring guide. Moves decisions to closure; the coalition held; the relationship survived the deal are outcomes any culture's machinery can produce and any assessor can verify. When you find a culturally thick verb in an anchor, commands the room, pushes back hard, visibly decisive, replace it with the function it was gesturing at, and check whether the role truly requires the style or only the fashion assumed it did. Occasionally it does require the style; a spokesperson role in New York has stylistic content. Then say so explicitly at scoping, as a role requirement with a reason, not silently, as everyone's assumption.

Split message from delivery in your notes. Two lanes on the page: what was said, meaning logic and evidence and structure and audience adaptation, and how it was delivered, meaning pace and volume and directness and warmth. Score the first against the anchors. Record the second as data about style fit, to be weighed only where the mandate genuinely prices it. The discipline sounds bureaucratic and takes one vertical line down a notebook page; it is the single cheapest de-biasing act in this chapter.

Psychometrics across borders

Chapter 22 governs instruments in general; here, the three rules that keep them honest internationally, because the failure mode is subtle: the score arrives with decimal places, and decimal places launder assumptions.

Equivalence, not translation. A test rendered into another language is not yet the same test; the technical standard is measurement invariance, evidence that the instrument means the same thing across the groups compared, and the international guidelines treat adaptation as a documented process, not a linguistic errand. Without invariance evidence, cross-country score comparisons are method artifacts wearing precision.

Response styles are real and cultural. Acquiescence and extreme responding vary systematically by country, demonstrated across EU samples and again across twenty-six nations, which means two identical executives can produce different profiles by answering style alone. The better publishers know this and say so.

Choose the norm group like it matters, because it does. Use the validated local-language form, not English-for-everyone; compare against the most relevant norm group available, not a global average by default; and treat country-mean differences as uninterpretable absent invariance evidence. The professional review of one major instrument put the applied problem in a sentence this book will simply borrow: which norm applies to an English applicant, for a role in a German multinational, based in Sweden? If your assessment provider cannot discuss that question fluently, the decimal places are decoration.

References across regions

The two-level method of Chapter 7 travels everywhere; what changes is the local weather, and four regions cover most of a cross-border firm's practice.

Germany: the written Arbeitszeugnis is a legal entitlement with a famous interpretive convention (the law prohibits coded language, the culture reads codes anyway), so treat the document as a formal record parsed by someone who knows the conventions, and let the disclosed oral second ring carry the evaluative weight. The United States: legal chill has thinned formal references to dates-and-title in much of the market; the structured oral format under documented consent recovers most of what policy removed, and the second ring, being personal rather than institutional, is less chilled than the HR letterhead. The Gulf: vouching runs through relationship networks, and the wasta literature is clear that careers and trust travel personal ties, which means the second ring is not an imported technique there; it is how the market already speaks. The method's job is to add discipline to it: consent-first, confidentiality absolute, and the crucial separation of network endorsement ("he is known to good families of business") from performance evidence ("I watched him rebuild the distributor network; here is what I saw"). Central Asia: the published evidence is thin and Kazakhstan-centric (clanism and connection effects are documented; executive-reference doctrine is not), so operate consent-first, weight episodes over reputation even more heavily than usual, and treat this book's own regional table as provisional, which is exactly what the underlying research says of itself.

Calibrating the assessors

The last adjustment points at the panel, because half the cross-cultural error budget sits on your side of the table. Before a cross-border search's interviews begin, brief every assessor on the specific style risks of the corridor in play, the known mis-scorings, by name: American self-enhancement read as substance; Germanic understatement read as low drive; Gulf relational choreography read as evasion; post-Soviet institutional narration read as lack of ownership. Then enforce the one control that the evidence says actually works: every score cites evidence — Chapter 16's rule, which is the cross-cultural defense too, because structure, not panel goodwill, is what eliminated similarity effects in the research. Diverse panels help, as legitimacy and accountability and a wider net for reading style, and they are not the fix. The citation discipline is the fix.

And police the second failure mode with the same energy as the first. Culture explains how evidence arrives; it never excuses evidence that doesn't. A candidate who, after the ladder and the briefing and every door this chapter opens, still cannot name a decision that was theirs, still cannot produce one episode of upward correction, still offers reputation where the bar demands episodes, has an evidence problem, not a cultural one. More doors. Same bar. The kindest thing about the mantra is that it protects the strong candidate from both injustices at once: the one that scores their grammar, and the one that stops expecting things of them.

One warning the whole chapter depends on

Region tables describe distributions, and you will never interview a distribution. The Munich candidate in front of you may be the most self-promoting person in Bavaria; the Riyadh candidate may be blunter than your Boston client. Cultural knowledge is a prior: it tells you which misreadings to guard against and which doors to have ready, and the person is the evidence. The moment the table starts predicting the individual instead of protecting them, it has become exactly the instrument of prejudice it was built to dismantle. Readers of the research corpus behind this book will recognize the shape of the rule: the candidate is not the profile, and not the cultural profile either.

Instrument: the two-layer protocol

One layer standardizes the construct and the bar; the other localizes the door. Printed for the process file:

Where the rules run out

The honesty paragraph. The evidence behind this chapter is strong exactly where the world's research money lives (the United States, Western Europe, the GLOBE map, psychometric adaptation) and thin where a firm like this one actually works some of its most interesting mandates: the Gulf evidence is solid on networks and sparse on executive assessment specifically; Central Asia's is thinner still and mostly Kazakh; the region table above is provisional and says so. Direct criterion studies of structured interviews compared across the target regions barely exist: the defensible claim is that structure remains the best base method everywhere studied, with culturally reviewed implementation, not that structure is culture-free. And the deepest limit is the warning above, which no protocol removes: the table protects individuals only while it describes populations, and the assessor who forgets the difference has automated the very bias the chapter exists to catch. Chapter 25's log, one more time, is the long-run judge, and cross-border placements, tracked honestly, are where a multi-hub firm can build regional assessment knowledge the literature does not yet contain.

Notes and sources

Evidence grades: [M] meta-analytic/systematic; [L] peer-reviewed primary; [S] credible practitioner/survey; [T] flagged synthesis.

• Self-presentation norms and rating effects; the transcript finding: Paulhus, Westlake, Calvez & Harms, "Self-Presentation Style in Job Interviews: The Role of Personality and Culture" (Journal of Applied Social Psychology, 2013). [L]
• Ten-country self-presentation survey (US high; Germany/Norway low): Sandal et al. (Journal of Cross-Cultural Psychology, 2014). [L]
• Universal vs. contingent leadership dimensions; cluster profiles: GLOBE program publications (House et al., 2004 and successors; cluster papers for Germanic and Latin Europe, the Anglo and Middle East clusters). [L]
• Power distance and upward voice: Kwon & Farndale (HRMR, 2020); power-distance and fear-of-authority findings. [L]
• Structured-interview transportability and cultural review recommendations: the SIOP cross-cultural interviewing white paper (Bourdage et al., 2021); expatriate-selection validity (Lievens et al., 2003; Mol et al., 2005). [S/L]
• Test adaptation and invariance: ITC Guidelines for Translating and Adapting Tests (2nd ed., 2017); measurement-invariance reviews. [S/M]
• Response styles by country: van Herk, Poortinga & Verhallen (2004); Harzing's 26-country study (2006). [L]
• Norm-group problem quotation: the British Psychological Society's OPQ32 review (2007). [S]
• German reference law and conventions: §109 GewO; chamber-of-commerce guidance. [S]
• Gulf networks: the wasta research literature; Saudi personal-networks studies. [L]
• Central Asia: Minbaeva & Muratbekova-Touron on clanism (2013); Hotho et al. on favoritism (2020) — Kazakhstan-centric, flagged. [L]
• The ladder, the two-lane notes, the corridor briefing, the mantra's formulation: this book's synthesis from the underlying research and practice. [T]